Cash flow, excess savings & the risks to consumer spending

Consumer spending is widely expected to be weak over the next few years given the significant squeeze from a higher cost of living and higher interest rates. However, spending could hold up for longer than thought given: (1) the squeeze on cash flows from interest rates is partly offset by higher interest income; (2) households could save less or even return to spending all they earn; and/or (3) more importantly, households could dissave, tapping the massive excess savings that were accumulated during the pandemic, which are worth about 20% of annual income in aggregate and are broadly distributed across the community relative to incomes.  

Consumer spending has been extraordinarily volatile during the pandemic, but slowed sharply at the end of last year, with annualised growth of 1½% in Q4 marking the smallest increase since the spell of weak growth in the year or so before COVID.  

Forecasting consumer spending is difficult, which is why the mistakes made in forecasting spending often account for most of the errors involved in forecasting the broader economy.

This is not surprising considering that spending accounts for about half of output and is clear from the close correlation between growth in consumer spending and the Reserve Bank’s forecast errors for real GDP growth over a one- and two-year horizon. 

At present, the market and the Reserve Bank are united in thinking that consumer spending is facing a long period of weak growth, assuming, of course, that interest rates are near a peak and that the economy avoids a recession.

This is captured in the Reserve Bank forecasting average annual growth in real spending of about 2% over the next two-and-a-half years of its forecast horizon.

To place this in perspective, the only time average growth has been weaker than this has been in the immediate aftermath of the global financial crisis and post-WW2 recessions.     

Constructing the bear case for consumer spending is straightforward, with a prolonged period of weak growth justified by:

1. The deep pessimism of consumers;
2. The squeeze on household income from a higher cost of living;
3. The cash-flow squeeze from the impact of higher interest rates on repayments of household debt; and
4. Lower asset prices, particularly house prices.

However, playing devil’s advocate, we have examined the risk that spending might hold up for longer than expected, focusing on the cash flow mechanism and the saving behaviour of households. 

This risk is important because if it is realised then additional demand would place upward pressure on underlying inflation and hence interest rates.  

We examined the cash flow-effect of higher interest rates because it is the part of the transmission mechanism of monetary policy that receives outsized attention from market participants.  

At face value, the story regarding cash flows is straightforward, pointing to interest rates placing a large squeeze on household incomes via higher repayments of household debt.

On our calculation, total payments of household debt – comprising scheduled payments of principal and interest on mortgages, consumer debt, and small business debt - should increase from a multi-decade low of 11% of household income in early 2022 to around 17%, assuming the cash rate peaks at around 4%.

This would almost match the 18½% peak reached in mid 2008, back when the global financial crisis was well under way and the cash rate peaked at over 7%.  

This substantial squeeze on household income will be muted by higher interest income, where deposits held by households have increased sharply during the pandemic.

On our estimation, interest income should increase from a recent multi-decade low of near zero as a share of household income to about 3% by next year (there is more uncertainty around the likely path for interest income related to how the higher cash rate is reflected in higher deposit rates).

Higher interest income will encourage some households to save more – which is another channel of the transmission mechanism of monetary policy – but some households will probably spend some of the money after several years of near-zero returns on deposits.  

Another way households can support spending is by making smaller - and potentially no further - additional mortgage payments into offset accounts and redraw facilities (most mortgages have redraw facilities, while about 40% of mortgages have offset accounts).

Households have been making sizeable additional payments into these types of accounts for more than a decade, with net additional payments equalling about 2% of household income at the end of last year. 

Additional payments fell as low as ½% of income prior to COVID, which suggests that households could make smaller extra payments if they want to support spending in the face of a higher cost of living and the cash-flow squeeze from higher interest rates.  

In a similar vein to potentially reducing additional mortgage payments, more significant support for spending could be provided by households changing their saving behaviour, either by continuing to save at a slower rate or even dissaving to tap the excess savings built up during the pandemic.

The scope for saving at a slower rate is shown by the best measure of households’ overall net cash flow, which is net lending, defined as gross saving less investment in physical assets (note that in comparison the widely-quoted household saving ratio is a narrower measure of cash flow).  

Cash flow on this metric briefly peaked at a record 21% of household income at the height of the pandemic and fell to about 1% at the end of last year.

Prior to the pandemic, net lending was broadly zero, which suggests that households could support spending by further saving at a slower rate and potentially returning overall cash flows to balance.  

A much larger support to spending could come from households dissaving and running down the excess savings built up during the pandemic, something already happening in the US.

Replicating work by the Federal Reserve, we estimate excess savings built up during COVID total about $0.3 trillion, or around 20% of annual household income.

The excess savings mainly reflect the extraordinary financial support provided by the government during the pandemic, as well as sharply lower consumer spending, particularly greatly reduced spending on services during the lockdowns.

In dollar terms, our analysis of cross-sectional data suggests that the distribution of excess savings is skewed towards older, wealthier households.

However, the distribution across age groups and incomes is much more even when excess savings are expressed as a share of each cohort’s income.  

This suggest scope for households across age groups and the income distribution to prop up spending for longer than currently anticipated.